The research projects encompassed by this annual report employ the methods of behavioral neurophysiology and behavioral biology to study the mechanisms of higher brain functions. The past year has seen the completion or publication of thirteen reports, reviews and other academic works, including four peer-reviewed research articles in archival journals, as well as several refereed reviews and encyclopedia articles. Most of our research involves studying the role of the frontal cortex in visually guided actions, including the mapping of symbols to actions. The selection and control of action is not always recognized as an important issue in psychiatry or in psychiatric research. However, what a person does and the basis on which actions are selected comprise fundamental features of human life. There is increasing recognition that diseases such as schizophrenia, attention deficit-hyperactivity disorder, and obsessive-compulsive disorder, may result from inappropriate selection and control of actions. The work on this project has shown that accurate and appropriate action depends upon the proper function of specific parts of the frontal lobes. Our research has thus been dedicated to understanding the functions of these areas of the cerebral cortex so that, eventually, physicians, therapists, and other health care specialists can develop improved treatments for people suffering those and other mental health disorders. Two kinds of visually guided actions have been studied in detail. In the first, a spatial visual cue serves either as the target of action or the subject of a perceptual report. In some our most important work to date, Lebedev, Douglas, Moody, and Wise (2001) discovered that prefrontal cortex neurons had activity that reflects perceptual reports about a visual cue and that they were intermixed with neurons reflecting the guidance of movements to the same cues. This finding, reported in the Journal of Neurophysiology, has important implications for theories that posit a separation of the neural pathways and networks underlying the use of visual inputs for perception, as opposed to the use of the same inputs as targets of movement. Lebedev and Wise (2002) are preparing a review of these and related findings for Behavioral and Cognitive Neuroscience Reviews, to be completed and published next fiscal year. Lebedev and Wise (2001) also found that activity in a particular part of frontal cortex, known as the premotor cortex, reflects the orientation of spatial attention. This finding is important, and surprising, because the premotor cortex is commonly considered a motor area of the cerebral cortex. The finding of nonmotor neural signals, therefore, indicates that the function of the premotor cortex is more general than commonly accepted, as first proposed in earlier work on this project (Vaadia et al., 1988). As part of the study leading to the Lebedev and Wise (2001) report, an entirely novel food-delivery device had to be invented, which was described in a methodological paper by Mitz, Boring, Wise, and Lebedev (2001). The second aspect of this project studies a different kind of visually guided behavior, one known as symbolic mapping. This higher-order problem-solving behavior is also known as conditional motor learning, conditional discrimination, visuomotor association and symbolic matching. In symbolic mapping, the choice of an action to be made depends on the behavioral context provided by a symbol. This behavior differs importantly from behaviors guided by spatial information or simply by rewards. In symbolic mapping, although the symbol may be in a particular place, that location has no relevance to the action or decision to be made on the basis of that symbol. Further, none of the symbols that form the basis for a decision have any more or less association with reinforcement or reward than any other symbol. Thus, subjects cannot solve the problems posed by using a primitive strategy, such as choosing a more positive symbol and avoiding a less positive one. Usually, visual cues serve as symbols and the decisions to be made are about the target of movements or the kind of movement. More generally, however, any information packet can be mapped, in this sense, to any other information packet. This is the basis for learning the meaning of most words, for learning to associate that meaning with the motor programs necessary to generate speech and language, and for the wide variety of symbol- and signal-guided behavior that underlies much of our behavioral flexibility. In studies related to this second kind of visually guided action, Brasted, Bussey, Murray and Wise (2001) reported that damage to parts of the hippocampal system impairs symbolic mapping when nonspatially differentiated responses are to be made in response to symbols. In this research project, an example of a nonspatial response is tapping a visual stimulus eight times. An alternative nonspatial response is maintaining steady contact with a stimulus for between four and eight seconds. The findings of Brasted et al., now submitted for publication and under review, challenge the prevailing doctrine that the function of the hippocampal system involves only spatial information processing. Our data, instead, support an idea proposed most formally by McClelland, MacNaughton and O Reilly (1995) and by Wise and Murray (1999), that the symbolic mapping functions of the hippocampal system are more general than commonly believed. This idea is consistent with the knowledge that hippocampal damage in amnesic patients severely disrupts language functions in addition to other forms of memory. Bussey, Wise and Murray (2001a, b) found that the information processing and transfer functions performed by the ventral and orbital prefrontal cortex are necessary for learning symbolic mappings at a normal rate, as is the interaction between those frontal areas and the inferotemporal cortex in the same hemisphere. In neurophysiological work on this topic, Chen, White and Wise (2001) studied frontal cortex activity during symbolic mapping and found that even though features of symbols may not be relevant to the decision to be made on the basis of those symbols, some of these features are nevertheless encoded in frontal cortex activity. In order to consider the importance of symbolic mapping to everyday life activities, Murray, Brasted and Wise (2001) reviewed its potential role in normal behavior as well as in the laboratory. In more general academic work, Moody and Wise (2001) reviewed connectionist contributions to population coding in the motor cortex, and Wise (2001) reviewed the neuronal mechanisms of movement selection, preparation, and the decision to act for a book on the Bereitschaftspotential. Wise (2002) also wrote the entry on motor cortex for the International Encyclopedia of the Social and Behavioral Sciences and co-authored the entry on motor control for the Encyclopedia of the Human Brain.